NeuroSift for Task-Aware Quality Assurance of Multimedia Data in Remote Parkinson Disease Assessment
Published in Journal of Medical Internet Research, 2026
We introduce NeuroSift, a task-aware, interpretable machine learning framework for assessing recording quality and task compliance in home-recorded multimedia data collected for remote Parkinson’s disease assessment. Covering three remotely collected tasks — finger-tapping, facial expression (smile), and speech (pangram utterance) — NeuroSift’s quality classification models achieved $\mathbf{77\text{–}90\%}$ ordinal accuracy against expert consensus, supporting more reliable downstream screening in unsupervised, at-home settings.
Recommended citation: Islam, M.S., Park, S., Ma, E.X., Adnan, T. and Hoque, E., 2026. NeuroSift for Task-Aware Quality Assurance of Multimedia Data in Remote Parkinson Disease Assessment. Journal of Medical Internet Research, 28, p.e91756.
Download Paper